STEVENSInstitute of Technology

E 321DESIGN V

Fall 2000

2

COURSE OVERVIEW

Materials are ubiquitous throughout our society. They are what we see as we go through our

lives each day, and they are what we use to improve the quality of our lives as time movesforward. Materials are so essential and so integrated into our lives that they are often simplytaken for granted by most everyone that uses them. Engineers, however, can not take materialsfor granted. Engineers must have some level of familiarity with this vast set of building blocksused in the construction of devices demanded by all aspects of our society.

Design V (E-321) is meant to provide experience within the Stevens Engineering core in the useand selection of materials for engineering applications. It achieves this in two principal ways.First is a series of twelve laboratory experiences which address various aspects of materialsprocessing, structure, and properties. These exercises sample the five classes of engineeringmaterials: metals, ceramics, polymers, semiconductors, and composites. The laboratories arebroken into four modules. Each module consists of three laboratory-based exercises. The fourDesign V modules [(I) Structural Materials ; (II) Functional Materials; (III) Materials Processing;and (IV) Materials Microstructure] are closely integrated with the four lecture modulesdeveloped in the separate course E-344 Materials Processing. E-321 and E-344 have beendeveloped so that students participate in the Design and Lecture modules on the same subject atthe same time.

The second and very much inter-related component of Design V centers on the selection ofmaterials. The effective design of any engineering structure requires the proper choice ofmaterials or sets of materials to meet specifications defining needed properties. The materials-selection component of Design V thus uses the experiences gained in E-321 laboratory and E-344lecture as a foundation from which to develop skills needed to formulate and answer relevantquestions associated with the selection of materials for engineering applications. Within thecontext of previous Stevens Design-Spine experiences on project management, teaming, ethics,engineering economy, and others, Design V introduces materials selection using the CambridgeEngineering Selector (CES) database. The CES database is one of several modern tools availableto practicing engineers to rapidly identify the specific materials which best satisfy propertyspecifications demanded by some broader engineering design process. Design projects aredeveloped within E-321 which present scenarios where sets of needed properties must be definedand materials must be selected in response to these properties. These scenarios force the need tooptimize a materials selection, since there is not necessarily a single correct answer but rather abest answer to a particular materials-design problem.

3

TABLE OF CONTENTS

PAGE

LABORATORY SAFETY 4

MODULE I – STRUCTURAL MATERIALS 6

• Tensile Properties of Materials 6

• Creep of Solder 11

• Cold Working and Annealing 16

MODULE II - FUNCTIONAL MATERIALS 21

• Characterization of a Silicon Wafer and Fabrication of a Solar Cell 21

• Solar Cell Evaluation 26

• Electrochemical Phenomena 30

MODULE III – MATERIALS PROCESSING 38

• Melting and Casting 38

• Powder Metallurgy 40

• The Preparation and Properties of Ceramic Ferrite Magnets 44

MODULE IV- MATERIALS MICROSTRUCTURE 47

• Heat Treatment of Steel 47

• Digital Microscopy 53

• X-Ray Energy Dispersive Analysis Of Materials 57

4

LABORATORY SAFETYGeneral Safety Tips

To practice safety is to be aware of your surroundings and potential hazards. Once youunderstand the chemicals and equipment used in the laboratory, the necessary safety precautionscan be taken. In general there are five principles of safety which should always be followed:

1) practice safety,

2) be concerned about the safety of others,

3) understand the hazards associated with your particular experiment,

4) know what to do in an emergency, and

5) report hazards or hazardous conditions.

1) Practice Safety

Often, we become complacent in our laboratory proceedings by taking short cuts whendealing with safety (i.e., not wearing safety glasses). This can be disastrous. It is important torealize that accidents can occur at any time. We must anticipate the worst and take necessaryprecautions. Below are some of basic safety precautions to be followed at all times:

a) Wear appropriate eye protection when working with potential eye hazards.

b) Use a hood for hazardous, volatile, and noxious chemicals.

c) Wear proper clothing (i.e., no shorts or open-toe shoes).

2) Be Concerned About The Safety of Others

Your concern for safety must also include your fellow colleges. Be aware of the people aroundyou. This can include something as simple as reminding a friend to wear safety glasses.

3) Understand The Hazards Associated With Your Particular Experiment

5

When working in a lab you must understand how everything works and what potential hazardsare associated with the equipment and chemicals used. Often the hazards are quite apparent andpracticing safety is just a matter of thinking about what could go wrong before it does go wrong.

4) Know What To Do In An Emergency

Just knowing the possible hazards is not sufficient. In an emergency, it is important to respondquickly. Take the time to familiarize yourself with your laboratory surroundings, (i.e., exits, fireextinguishers, fire alarms, eyewash stations, etc.).

5) Report Hazards or Hazardous Conditions

Often accidents can be prevented if a hazard is identified. If you have witnessed or notice apossible hazard, you must immediately notify a teaching assistant. This could include falteringwiring of electronics or a simple suggestion on how to conduct the lab in a safer manner.

Specific Hazards

Each laboratory module has associated hazards. Therefore it is important to be aware of suchdangers and the corresponding precautions. At this time, safety precautions for the four sectionsof Design V are discussed below.

MODULE I – STRUCTURAL MATERIALS

In this laboratory, the biggest hazard is the rolling machine. This machine is used toflatten metal billets into thin strips by two rolls. It is important never to use your fingers tointroduce the billet into the rolls. Tongs are provided for this job.

MODULE II - FUNCTIONAL MATERIALS

In this section, a very dangerous chemical, hydrofluoric acid (HF), is used to as an etchant. Boththe vapor and liquid solution are extremely toxic. All forms of HF are easily absorbed throughthe skin and can cause long-term, excruciating burns without any immediate indication ofexposure. Therefore there are several precautions which must be taken:

1. Always use HF under a hood. This also applies to the electroless nickel solutions.

2. When working with HF, wear chemical gloves, apron and safety glasses. All safetyequipment is provided in class.

3. Proper clothing is important in this lab. Try to minimize the amount of exposed skin.

6

If exposure to HF does occur, quickly flush the area with cool water until any whitening of thetissue has subsided. Note that simple flushing of the skin does not remove all traces of HF deepin the tissue. Continue to cover the area with wet cloths after flushing and seek immediatemedical attention. Other potential hazards in this lab are the furnaces and the soldering irons.Protective eye-wear should always be worn when operating this equipment.

MODULE III – MATERIALS PROCESSING

In this group of lab experiments, a very powerful furnace is used to melt metals. The meltingfurnace operates at 450 V and about 5 to 10 kW of power. Although the machinery is wellshielded it is important never to touch any of the leads while in operation. Other precautionsshould be taken in regard to the elevated temperatures if the molten metal. The following is abrief list of necessary precautions:

1. Never touch the furnace until the teaching assistant has turned off the power.

2. Always wear safety glasses when working near the furnace (i.e., pouring orskimming oxides from the surface).

3. Make sure that the molds are dry prior to pouring the molten metal.

4. Wear proper clothing and fire resistant aprons provided in class.

The small furnaces used in the lab are similar to those in the Semiconductor lab andrequire the same precautions. Another potential hazard is the band-saw used to cut metalsamples. Remember never to place hands near the blade while the saw is on. For example, whena sample is sectioned by the blade, always turn the saw off before grabbing the cut piece.

MODULE IV- MATERIALS MICROSTRUCTURE

In this group of lab experiments you will be working with a x-ray device that emits lowlevel radiation. Therefore there are some precautions that must be abided by when dealing withthe equipment.

1. Under no circumstances should any part of the body be placed directly in primary x-ray beams.

2. Whenever possible, place the shutter in the closed position to impede the primary x-ray beam.

3. Never align samples with the eye in such a way as to be exposed to the primary beam.

7

MODULE I – STRUCTURAL MATERIALS

I.1 THE TENSILE PROPERTIES OF MATERIALS

Objective

This experiment introduces the stress - strain behavior of a variety of materials (carbonsteel, polymers, aluminum alloys, etc.) and the tensile test. By conducting tensile tests onsamples in accordance with the ASTM standard A-370 (www.astm.org), the modulus ofelasticity, yield strength, percent elongation, strain hardening coefficient, strength coefficientsand toughness will be determined.

Supplemental Reading

SMITH6.2 - STRESS AND STRAIN IN METALS6.3 - THE TENSILE TEST6.9 - FRACTURE OF METALS (DUCTILE/BRITTLE)

OHRING7.2 - ELASTIC BEHAVIOR7.3.1 - 7.3.3 - PLASTIC DEFORMATION OF METALS7.4.1; 7.4.2 - DISLOCATIONS

see also www.instron.com

Background

The tensile test is perhaps the most widely used mechanical test. Tensile tests providebasic design information on the strength of materials which is often specified in engineering codes.In a tensile test the specimen is fastened in a pair of grips, one connected to a load cell and theother to a movable cross head. By moving the cross head, the specimen is subjected to anincreasing uniaxial load. The load (P) and specimen elongation (∆l) are measured and constitutethe essential quantities displayed on a stress - strain curve. There are two kinds of stress - straincurves described below:

1. Engineering Stress - Strain Curve is a plot of σe (engineering stress) vs. εE

(engineering strain).

σe = Pi/Ao … (1)

where Pi is the instantaneous load and Ao is the sample’s original cross-sectional area.

8

εE = ∆l/lo … (2)

where ∆l is the change in length and lo is the original sample length.

2. True Stress - Strain Curve is a plot of σT (true stress) vs. εT (true strain).

σT = Pi/Ai … (3)

where Pi is the instantaneous load and Ai is the sample’s instantaneous cross-sectional area.

lo+∆l

εT = ∫ dl/l = ln ((lo + ∆l)/lo) = ln (1 + εE ) …(4)

lo

Ai= Ao⋅ exp(-εT) … (5)

where Ai is the instantaneous area and Ao is the original area.

The stress-strain behavior for most materials can be described by two regions, elastic andplastic. In the elastic region, small strains are linearly related to the corresponding stress, seeequation (6).

σe = E⋅ ε E …(6)

where E is the modulus of elasticity.

In the plastic regime, the stress is related to the corresponding strain by equation (7):

σT = K⋅ ε Tn …(7)

where K is the strength coefficient and n is the strain hardening coefficient.

With the above relations, it is possible to determine quantities such as modulus ofelasticity, yield strength, percent elongation, offset yield strength, modulus of resilience,toughness, etc. Some of these will be obtained in this experiment.

9

Procedure

1. Necessary Measurements

1.1 The cross-sectional area of each sample will be measured with a hand-held micrometer and recorded in your lab notebook. Mark a distance of 2 inches, centered along the length of each specimen.

2. Conducting the Tensile Test

2.1 The samples are placed in the lever-actuated wedge-shaped grips used in conjunction with the Instron tensile machine (www.instron.com).

2.2 Note the crosshead speed before starting the machine, then begin the test. The Instronmachine will plot the applied load (P) vs. time. Thus the strain can be obtained knowing thechart speed (1 inch per minute), the gauge length and the crosshead speed (strain rate). With thisinformation along with the original cross-sectional area of each specimen, plot the engineeringand true stress- strain curves for each material. A schematic example of engineering stress plottedagainst engineering strain is given in the figure below for two hypothetical alloys.

σ (MPa)

ε (m/m)0.01 0.030.02 0.04

100

500

400

300

200

0.002

** *YP

UTS

BPAlloy X

Alloy Y

10

3. Calculations

3.1 From each stress-strain curve, determine the modulus of elasticity (E), ultimate tensile strength (σUTS), yield strength (σYS) and the percent elongation at σUTS and

σYS. Be sure to specify the units.

3.2 Graphically determine K and n for all materials. Note the following useful relation:

ln(σT) = ln(K) + n⋅ ln(εT) …(8)

Plot ln(σT) vs. ln(εT), from which you can determine n and K.

Questions

1. What is the significance of n insofar as deep-drawing or forming operations are concerned?

2. Explain the difference in behavior between the polymer specimens.

3. Why do the polymer specimens develop a white opaque appearance upon plastic loading?

4. Explain the observed differences in the stress-strain behavior of the annealed and cold-worked steel samples.

11

TENSILE TESTING

Sample Area (in2)

σYS

(Psi) εYS σUTS

(Psi) εUTS E

(Psi) ∆l (in)

K (Psi)

n

Alalloy

Steel

LDPE

HDPE

12

I.2 CREEP OF SOLDER

Objective

This experiment introduces a time-and-temperature dependent deformation mechanism calledcreep. A low melting point Pb-Sn alloy will be used to study the creep phenomenon as well asderive useful creep parameters for design and materials selection.

Supplemental Reading

SMITH 6.11 - CREEP AND STRESS RUPTURE OF METALS

OHRING 7.7.3 - CREEP

Background

Creep of materials is a phenomenon common to the glasses employed in medieval stained-glass windows as well as to the superalloy turbine blades in modern jet aircraft. Under the actionof a constant load at "relatively elevated" temperatures, materials are observed to stretch withtime. When the load is first applied, an instantaneous elastic elongation occurs. The strain ratethen decreases rapidly with time in the first (primary) or transient creep stage. With time thetransient state gives way to the second (secondary) or steady-state creep stage which ischaracterized by a constant minimum rate of stretching as the strain hardening is balanced byrecovery (or recrystallization). This state is of practical importance and much creep design workis concerned with characterizing the steady-state deformation. Theoretically the steady-statecreep strain state ε is given by:

ε = A⋅ σ n⋅ e-Q/RT …(1)

where σ is the stress.Q is the activation energy (J/mol)T is temperature (K)R is the gas constant (J/mol⋅ K)A and n are constants.

Finally, during the third or tertiary creep stage the strain rate accelerates and ultimately failureoccurs. Depending on the particular values of A, n and most importantly Q, materials creep atvery varied rates. For example, solder creeps at room temperatures under modest stress whileceramic materials creep only at very high temperatures under considerable loading. The values ofσ and Τ involved can accelerate creep to occur in a matter of minutes or extend it to a period ofyears.

13

Procedure

1. The Creep Frame and Experimental Setup

1.1 The creep experiment is conducted on a specially designed frame (see schematic below).

To incorporate temperature as a variable, the test temperature will be varied by passing currentthrough the wire (resistive heating) and controlled with a thermocouple device. A constant loadis applied by hanging weights to the wire rope connected to the specimen. Elongation ismeasured by a dial gage indicator reading accurate to 0.001 inch.

1.2 The specimen wire to be tested should be fastened in the movable grip first, then drawntaut through the fixed grip and fastened there such that the dial gage just begins to show a smalldeflection. Where possible, use a double thickness of wire in the grip.

1.3 Set the dial gage to zero, add a load and note the deflection as a function of time.Take an elongation reading every 15 seconds for 8 minutes. Record the data in the tablesprovided in this packet.

1.4 A total of eight tests will be conducted, four tests at a constant temperature and varying loads and four at a constant load and varying temperatures.

2. Calculations

2.1 Determine both n and Q graphically by plotting ln(ε) vs. ln(σ) and ln(ε) vs. 1/T, respectively. Note the following useful equations derived from equation (1).

ln(ε) = ln(A)⋅ (-Q/R⋅ T) + n ⋅ ln(σ) (at constant T) …(2)

Frame

Dial Gauge

Weight

Solder wire (1/16”)

14

ln(ε) = ln(A)⋅ n⋅ ln(σ) - (Q/(R)(1/T) (at constant σ) …(3)

Equations (2) and (3) represent a straight line.

Questions

1. When current is passed through the wire, it elongates due to thermal expansion. (Consider the extent of this expansion in correcting your results.) How would you use this equipment to measure the coefficient of thermal expansion of solder?

2. What are some of the microscopic damage mechanisms and processes occurring within the metal during creep?

3. What is the engineering importance of the steady-state creep rate and why is the secondary stage the most important regime?

15

CONSTANT LOAD

1. Record temperature at start Tx= column heading2. Record Dial gauge reading adjacent to the time interval for the temp. being used3. Maximum temperature is 60 C4. From this table you will be able to produce TWO GRAPHS

a) STRAIN ( ε ) vs TIME for the four temperatures ( stress is constant )

b) CREEP RATE ( ln ε. ) vs 1/ TEMPERATURE ( from four slopes part a )

TIME T1= T2= T3= T4=

15 sec30451:00 min1:151:301:452:002:152:302:453:003:153:303:454:004:154:304:455:005:155:305:456:00

16

CONSTANT TEMPERATURE

1. Record load pounds at start Lx= column heading and above in PSI2. PSI = (applied load)/ 3.06 x 10 -3 in2

3. Record Dial gauge reading adjacent to the time interval for the load being used4. From this table you will be able to produce TWO GRAPHS

a) STRAIN ( ε ) vs TIME for the four loads ( temp. is constant )

b) CREEP RATE ( ln ε. ) vs ln σ ( from four slopes part a )

TIME L1= L2= L3= L4=

15 sec30451:00 min1:151:301:452:002:152:302:453:003:153:303:454:004:154:304:455:005:155:305:456:00

17

I.3 COLD WORKING AND ANNEALING

Objective

This lab experiment quantitatively demonstrates the effects of cold rolling and annealing on severalaluminum-based materials. The materials used are cast billets of the following compositions: pureAl, Al - 2wt% Si, Al - 4wt% Si and Al - 6wt% Si. The effect of cold rolling and annealing will bequantified by observing changes in the material’s hardness.

Supplemental Reading

SMITH 6.1 - METAL PROCESSING (ESPECIALLY PP 251-256)6.6 - PLASTIC DEFORMATION OF METALS6.7 - SOLID-SOLUTION STRENGTHENING MECHANISMS6.8 - RECOVERY AND RECRYSTALLIZATION

OHRING 8.3.4 - ROLLING9.4 - MECHANICAL WORKING AND RECRYSTALLIZATION9.5 - STRENGTHENING NONFERROUS METALS

Background

Cold working includes the widely employed rolling, wire drawing, sheet metal forming, stampingand machining operations. When a metal is cold worked, it deforms plastically and hardens as aresult of complex increases in dislocation density and the generation of impediments (e.g., latticestacking faults, dislocation pile-ups, jogs, etc.) to dislocation motion. In order to produceadditional deformation, increasingly greater loads are required which reflect an increase in metalstrain due to work hardening. Generally the strength(S) varies with strain (e) as S = Ken where K isa constant, and n is the strain-hardening coefficient. In the laboratory, the cast ingots will be rolledto varying amounts of reduction and the extent of hardening or strengthening will be evaluatedthrough hardness measurements. The hardness test essentially measures the penetration depth ofeither a steel ball or a diamond indentor into the metal surface due to an applied load. The harderthe metal, the less the penetration and the higher the hardness number. While hardness is a measureof a materials resistance to penetration, and is directly proportional to tensile strength for somemetals (steel), it does not provide information on ductility.

When deformed metals are annealed at sufficiently high temperatures where diffusion rates areappreciable, the net effect of the elevated temperature is to undo the effects of the previousdeformation. Both point and line defects are reduced in number and spatially altered such that thestrain and surface-free energies are reduced thereby softening the material. Eventually new strain-

18

free grains with relatively low dislocation density are produced (at high annealing temperatures).Annealing treatments frequently follow cold-working steps to increase ductility and decreasematerial hardness and susceptibility to cracking. Several stages in softening have been identified -recovery, recrystallization and grain growth. In this lab, you will observe the onset of thissoftening process through the use of a hardness test.

wt% Si

T (

o C)

577 oC

660

.37

oC

19

Procedure

1. Cold Rolling and Hardness Test

1.1 Take several hardness measurements for the pure Al and three Al-Si alloy specimens provided in class. The hardness test and scale used in this experiment will be a Rockwell H scale. Make sure the specimen lies flat on the hardness anvil. Otherwise the specimen bow will distort readings.

1.2 Measure the initial thickness of each sample before and after rolling.

1.3 Each sample will be cold rolled to approximately 75% reduction in cross-sectional area. After every 15% reduction in area, measure the specimen hardness.

1.4 Plot hardness vs. percent reduction of each sample. To calculate the exact percent reduction, use the following equation.

% Reduction = (Ao - Ai )/Ao

where Ao is the original cross-sectional area and Ai is the instantaneous area

2. Annealing and Recrystallization

2.1 After rolling, anneal the specimens for about 20 minutes at 350°C. Cool the specimens and remeasure the hardness.

2.2 Note the effect of annealing on each sample.

Questions

1. What microstructural changes occur during cold rolling, annealing and solution hardening?

2. How did the hardening coefficient vary between the different aluminum-based materials, and does this make sense?

3. Why were the materials only reduced in area by 15% for each roll? Why not 50 or 75% for each roll?

20

COLD WORKING TABLE

Codes for Materials:

1. P = Pure Aluminum

2. R = 2% Si - Al

3. S = 4% Si -Al

4. T = 6% Si- Al

Sample E0 E1 E2

T0 H0 T1 H1 T2 H2

PRST

Sample

E3 E4 E5 E6

P T3 H3 T4 H4 T5 H5 T6 H6

RST

21

MODULE II - FUNCTIONAL MATERIALS

II.1A CHARACTERIZATION OF A SILICON WAFER

This project consists of several parts. Parts II.1A and II.1B will be done during onelaboratory meeting.

Objective

The purpose of this lab is to characterize the semiconductor type and resistivity of asingle-crystal silicon wafer by utilizing the thermoelectric effect and a four-point probetechnique, respectively. Once the material has been characterized, each wafer is subjected to adoping procedure, the first step in processing a p-n junction device or diode.

Supplemental Reading

SMITH 5.1 - OHM’S LAW (PP. 187-189)5.3 - INTRINSIC SEMICONDUCTORS5.4 - EXTRINSIC SEMICONDUCTORS

OHRING 12.2.1; 12.2.2 - INTRINSIC/EXTRINSIC SEMICONDUCTORS12.3.3 - THE SEEBECK EFFECT

Introduction

Silicon semiconductor technology spans applications ranging from microelectronics andcomputers to solar photovoltaic cells (solar cells). The starting materials for both of theseapplications are silicon wafers sliced from large grown single crystals of silicon. Subsequentapplications and processing depend to a large extent on the properties of the original Si; theelectrical type (p-type or n-type) of the intentionally added dopants and their concentration.The crystal perfection of the Si wafer is also important because microelectronics applicationsrequire defect-free, highly polished wafers of very uniform dopant concentrations. For solar cellsand large discrete devices, (i.e., diodes and transistors) the perfection and uniformityrequirements are less critical. In order to obtain high yields of acceptable devices, it is importantto characterize the initial silicon wafers. It is essential to use experimental methods which arequick, reliable and nondestructive. Furthermore, it is desirable that measurements be madeutilizing only ONE face of the wafer. Two such widely used methods for satisfying the aboverequirements have been developed and will be employed in this laboratory to determine theresistivity and wafer type. Below are brief descriptions of each technique.

22

1. Silicon Wafer Typing

[THERMOELECTRIC EFFECT]

In this technique, both a heated and aroom- temperature probe contact anunknown wafer. The hot probe heats thesample producing a slight voltage that canbe detected by connecting a voltmeteracross the probes.This is known as the Seebeck Effect. Thehot probe is positive with respect to thecooler probe for n-type Si and negativefor p-type Si.

2. Wafer (Sheet) Resistance

[FOUR-POINT PROBE METHOD]

In this technique, four precisely alignedprobes make contact with the waferspecimen which can either be doped orundoped. A known DC current (I=6 mA)is passed through the outer two probeswhile a voltage is measured across theinner two probes. The probes are alignedin a linear arrangement. k = 4.53 (probegeometry constant). The thickness of thewafer, t = 0.02 cm.

Digital Voltmeter

SolderingIron

Probe

Wafer

2 3

I

V

1 2 3 4

23

The sheet resistance (Rs in ohms per square) is calculated by:

Rs=[(voltage in mV)(k)]/I …(1)

and the sample resistivity (ρ in ohm⋅ cm) is related by: ρ= (Rs)(t)

k is a constant which depends on the probe geometry. The geometry used in the Design V labuses a geometry for which k = 4.53. The wafer thickness is t = 0.02 cm.

In a diode, t may be taken as the junction depth. The constant 4.53 arises from thesolution to the electrostatic problem associated with the spreading of current from the twooutside probes and depends on specimen and probe geometry, as well as wafer thickness. Rs isknown as the sheet resistance and is a commonly utilized term in thin film and semiconductortechnology. Rs is normally a measure of dopant concentration.

Procedure

1. The Characterization of the Silicon Wafer

1.1 Using the Hot Probe Method, determine the wafer type employing a soldering iron as thehot probe and a regular cold wire probe. Clip the positive lead of the digital voltmeter (DVM) tothe heated iron tip. The cold probe is then connected to the negative lead of the digital voltmeter.

1.2 Simultaneously touch both probes to the wafer surface (without letting the two probestouch). Note the polarity of the voltage. The voltage will enable you to determine if you have ap-type or n-type wafer.

2. Etching Procedure

2.1 In this part of the experiment, you will use a 10% solution of hydrofluoric acid (HF) toetch away the SiO2 oxide layer. The oxide is an insulator and must be removed prior tomeasuring the sheet resistance.

2.2 HF IS AN EXTREMELY HAZARDOUS ACID. IT CAUSES SEVERE BURNSWHICH MAY NOT BE IMMEDIATELY PAINFUL OR VISIBLE. AVOID CONTACTWITH SKIN, EYES, AND CLOTHING. AVOID BREATHING THE VAPOR. (HF has agreat affinity for bone calcium and will cause painful burns, particularly underfingernails). IN CASE OF CONTACT WITH SKIN, WASH IMMEDIATELY WITHPLENTY OF WATER FOR 15 MINUTES. HF IS TO BE HANDLED ONLY IN THEHOOD WHILE WEARING PROTECTIVE GOGGLES AND GLOVES.

24

2.3 Place the wafer in a blue plastic basket stored in the hood and lower the basket into the HFsolution for 5 minutes. After the first 2 minutes of etching, begin to agitate the solution for theremaining 3 minutes.

2.4 Take basket out and rinse in tap water.

2.5 Remove the wafers from the basket with tweezers and rinse with acetone.

2.6 DO NOT HANDLE WAFER WITH YOUR BARE HANDS!

2.7 Dry the wafers with the hot air blower.3. Wafer Sheet Resistivity Measurement Employing the Four-Point Probe Test

3.1 Swing the arm over so that the probe head gently makes contact with the wafer surface.

3.2 Keep the probe head in contact with the wafer surface just long enough to record avoltage reading. Take 3 readings on both sides of the wafer, making sure to move the wafer to anew position under the probes for each reading.

3.3 Calculate Rs and r for all readings. Take the average of the resistivity measurements ofboth sides of the wafer combined.

3.4 Using the graph of the Dopant Density as a function of ρ provided in this section of thelab notes, estimate the dopant density of your wafer.

Questions

1. For n-type Si, why is the hot probe positive with respect to the cooler probe and why is itnegative for a p-type wafer?

25

II.1 B FABRICATION OF A SOLAR CELL

Objective

During this part of the laboratory, the silicon wafers will be further processed into ashallow junction diode or solar cell. Dopants will be diffused into the solar cell. Thecorresponding junction depth will be determined from Fick’s second law of diffusion. Furtherprocessing steps such as masking and the application of contacts will also be covered.

Supplemental Reading

SMITH 4.6 - ATOMIC DIFFUSION IN SOLIDS ESPECIALLY NON-STEADY-STATE DIFFUSION (P. 163)4.7 - INDUSTRIAL APPLICATIONS OF DIFFUSION ESPECIALLY EXAMPLE PROBLEM 4.94.8 - EFFECT OF TEMPERATURE ON DIFFUSION5.6 - MICROELECTRONICS (PP. 228-233)

OHRING 6.2.1 - THE MATHEMATICS OF DIFFUSION (SEE EXAMPLE 6-2)6.3.2 - EFFECT OF TEMPERATURE12.5.3 - DOPING

Background

The introduction of dopant material into a wafer is based on a diffusion process.Equation (1) is a simple solution to Fick’s Second Law of Diffusion (∂C/∂t = D∂2C/∂x2) assumingthat diffusion occurs in one direction and that the diffusion coefficient (D) is not a function ofcomposition.

(Co-C)/Co = erf (x/[(2)(sqrt Dt)] …(1)

where Co is the concentration of dopant (1020 atoms/cm3) C is the concentration of P at the P-N junction D is the diffusivity (10-13 cm2/s) t is the time (3600 sec.)

x is the junction depth.

It is the diffusion of this dopant into the silicon wafer that creates the p-n junction.

26

Once the p-n junction is completed, the ohmic contacts must be applied to the wafersurfaces. This is often a difficult and complex step in semiconductor device fabrication. Toincrease the current collection efficiency of a solar cell, the top electrode should ideally be in theform of a grid. For simplicity, our cells will not have a grid contact but rather one small area ofcontact. The placement of the electrical contacts are controlled by a masking technique. Anelectroless nickel plating solution is used to deposit a contact on the silicon surface. The maskingtechnique controls the deposition, allowing the plating to occur only on exposed regions of thewafer surface. Electrical contacts such as a copper wire can then be soldered to the depositednickel region.

Procedure - Diffusion of Dopant Into Silicon

1. Doping Procedure

1.1 Depending on your wafer type, apply either a phosphorus or boron-rich liquid dopantsource to ONE side of the wafer. If your wafer is p-type spin on the phosphorus dopant, andif it's n-type, spin on the boron dopant.

1.2 Place the wafer on the photoresist spinner and apply the vacuum.

1.3 Press START to spin the wafer and apply 3 drops of dopant to the wafer surface. Allowthe wafer to spin for 30 seconds then press STOP.

1.4 Bake for 5 minutes at 250°C in the furnace.

1.5 After wafer is cool, trace the outline.

2. Dopant In-Diffusion

2.1 The wafers are placed in the furnace for 1 hour at 1000°C to drive the dopant into the wafer surface, forming the p-n junction.

2.2 Because of the high temperature and long time, the wafer surface has been oxidized, andrequires etching. Repeat the Etching Procedure from week 1 using an etching time of 20minutes to remove the oxide layer.

3. Calculations

3.1 During the etching procedure, take the time to determine the junction depth (X) usingequation (1). An error function table is provided below.

27

z erf(z) z erf(z) z erf(z)

0.000 0.0000 0.550 0.5633 1.300 0.93400.025 0.0282 0.600 0.6039 1.400 0.95230.050 0.0564 0.650 0.6420 1.500 0.96610.100 0.1125 0.700 0.6778 1.600 0.97630.150 0.1680 0.750 0.7112 1.700 0.98380.200 0.2227 0.800 0.7421 1.800 0.98910.250 0.2763 0.850 0.7707 1.900 0.99280.300 0.3286 0.900 0.7970 2.000 0.99530.350 0.3794 0.950 0.8209 2.200 0.99810.400 0.4284 1.000 0.8427 2.400 0.99930.450 0.4755 1.100 0.8802 2.600 0.99980.500 0.5205 1.200 0.9103 2.800 0.9999

4. Repeat the Four-Point Probe Test for Both Wafer Sides

5. Masking Procedure and Contact Deposition

5.1 Place a piece of tape on both sides of the wafer. Each piece of tape should have asquare hole exposing the wafer. On the n side, the hole should be off centered and small. Theexposed area on the p side should be larger. See illustration on the following page.

5.2 Place the wafer into the electroless nickel solution for 5 minutes. Note that the pH mustbe maintained above 7 and the temperature between 80 and 85°C. The pH is controlled byadding Ammonia to the solution.

Questions

1. What is the purpose of off centering the ohmic contacts on the surface of the solar cell?

2. Why are the four-point probe measurements taken during this week different than themeasurements taken during the first part of this lab?

Wafer (n side) (unexposed)

Exposed area (n side)

Exposed area (p side)

28

II.2 SOLAR CELL EVALUATION

Objective

This lab evaluates the current-voltage characteristics of the solar cell devices fabricatedpreviously. Each student will assess the efficiency and quality of their device by constructing anelectrical circuit for the measurement of the I-V characteristics.

Supplemental Reading

SMITH 5.5 - SEMICONDUCTOR DEVICES

OHRING 12.3.2.1 - SEMICONDUCTOR JUNCTIONS12.4.1 - SEMICONDUCTOR DIODE12.4.2 - METAL/SEMICONDUCTOR DIODE

Background

The solar cell you have fabricated isessentially a shallow junction diodeand has the well- known current-voltage characteristics typical ofdiodes. For example, in the forwardbias condition, the current risesexponentially with voltage while inthe reverse bias, a small reversecurrent flows. When light shines onthe junction with zero appliedvoltage, a small fourth quadrant ofthe I-V characteristics (positivevoltage-negative current) appears,representing the production ofelectrical energy which can beutilized in an external circuit. Thesize of this fourth quadrant is ameasure of the quality and efficiencyof the cell. Below is a schematic ofan I-V curve showing dark andilluminated junctions given by thedotted and solid lines, respectively.

V

I

(Vmax, Imax)

Quadrant IV

29

Procedure

1. Construction of the Circuit Used To Evaluate the Solar Cell

Build a circuit using the solar cell you have constructed according to diagram below.

2. Dark I-V Characteristics of Solar Cell

2.1 Take measurements of Vapp ranging from 12 (forward bias) and 12 (reverse bias). Recordthe corresponding currents and plot I vs. V1.

3. Light I-V Characteristics of Solar Cell

3.1 Measure the open circuit (the voltage when ISC=0) and the short circuit current (the currentwhen VOC=0).

3.2 Set V1 equal to zero and then record the voltage readings until VC goes to zero. Record atleast 6 to 8 data points and plot I vs. V1.

4. Calculations

4.1 From the I-V plots, calculate the fill factor (FF)

FF = [(Vmax)⋅ (Imax)]/[(VOC)(ISC)

Questions

1. What is the significance of the fill factor and how does your cell compare to the industrialvalue of 0.8? Give possible explanations for any differences in device quality.

+ -

Vapp

V1

Vc

100 Ω

P

I

N

30

II.3 ELECTROCHEMICAL PHENOMENA

Objective

The objective of this laboratory is examine aspects of corrosion related to engineeringmaterials, primarily metals. The laboratory will include:

1. Measurement of the electrode potentials of metals/alloys in simulated seawater.

2. Development of a galvanic series based on the measured electrode potentials.

3. Prediction of galvanic corrosion based on the position of metals in the galvanic series

4. Prediction of the possibility for cathodic protection based on the position of metals inthe galvanic series.

5. Demonstration of electroplating

Supplemental Reading

SMITH SECTIONS 12.1, 12.2, 12.3, 12.7 OHRING SECTIONS 10.1, 10.2

Background

Corrosion is the degradation of a material due to contact with the environment. Corrosion oftenoccurs in a liquid environment. We typically refer to this as aqueous or chemical corrosion.Corrosion can also occur from exposure to a corrosive gas atmosphere. This form of corrocionusually occurs at relatively high temperatures and is thus usually referred to as high temperaturecorrosion.

Corrosion, whether in liquid or gas, results from an electrochemical process. The primaryexception to this is the environmental degradation of polymers, which normally occurs by a non-electrochemical process such as the breaking of bonds by chemical means or by the absorbedenergy from ultra-violet light for example. In the other modes of corrosion that follow anelectrochemical pathway, there exists the transfer of electrons between two (or more) processes,which can both be occurring on the same material or be physically separated.For example, if a metal is corroding in an aqueous environment, the metal atoms lose valenceelectrons and become ions in the corroding liquid, thus the metal is removed (corroded) into theliquid environment. This corrosion process, which is known as an oxidation reaction, can only besustained if another reaction consumes those electrons that were released by the dissolving metalions. The reaction to consume the electrons is known as a reduction reaction. Hence corrosioninvolves coupled oxidation-reduction with electron transfer between the two reactions. This

31

knowledge actually provides one means to control corrosion. If the oxidation and reductionreactions are physically separated, an electrical insulator placed between the two locations willblock the electron transfer and hence stop the corrosion process from proceeding. This is used toprevent what is known as galvanic corrosion between dissimilar metals, such as the copper skinand steel supports in the Statue of Liberty.

Components of an Electrochemical Cell

Anode: The anode gives up electrons to the external circuit and corrodes.

Cathode: The cathode receives electrons from the external circuit and donates electronsto the electrolyte by means of a chemical reaction. Ions that combine with the electronsproduce a byproduct at the cathode.

Contact: The anode and cathode must be electrically connected permitting electron flow.

Electrolyte: The electrolyte is the liquid or gas in contact with both the anode andcathode. It provides the means by which the metallic ions leave the anode surface andassures that ions move to the cathode to accept the electrons.

Reactions of an Electrochemical Cell

Oxidation

Oxidation is a metal anode reaction by which metal atoms are ionized. The metalions enter the electrolytic solution while the electrons leave the anode through theelectrical connection:

M → Mn+ + ne-

Because metal ions leave the anode, the anode corrodes.

Reduction

1. The Electroplating Reaction: Reduction reactions are the reverse of the anode reactions -Electroplating:

Mn+ + ne- → M

The metal ions, either intentionally added to the electrolyte or formed by the anodereaction, combine with electrons at the cathode. The metal then plates out and covers thecathode surface. In many cases plating does not actually occur. A byproduct is formed atthe cathode instead.

32

2. The Hydrogen Electrode: In oxygen-free liquids, such as HCl or stagnant water, hydrogengas may be evolved at the cathode:

2H+ + 2e - → H2

A classic example of this is the attack on zinc by hydrochloric acid. When zinc is placed inhydrochloric acid it dissolves forming zinc chloride with the release of hydrogen gas. Duringmetallic corrosion, the rate of oxidation equals the rate of reduction in terms of electronproduction and consumption.

Zn → ZnCl2+ 2e - or Zn → Zn2+ + 2e - (Anode reaction)

2HCl + 2e- → H2 or 2H+ + 2e - → H2 (Cathode reaction)

Zn + 2HCl → ZnCl2 + H2 or Zn + 2H+ → Zn2+ + H2 (Overall reaction)

2. The Oxygen Electrode: In aerated water oxygen is available to the cathode and hydroxylions form:

1/2O2 + H2O + 2e- → 2(OH)-

The oxygen electrode enriches the electrolyte in OH- ions. These ions react withpositively charged metallic ions and generally produce a solid product. In the case of therusting of iron:

Fe → Fe2+ + 2e - (Anode reaction)

1/2O2 + H2O + 2e- → 2(OH)- (Cathode reactions)

Fe2+ + 2(OH)- → Fe(OH)2

Fe + 1/2O2 + H2O → Fe(OH)2 (Overall reaction)

The reaction continues to produce rust.

Fe(OH) 2 + 1/2O2 + H2O → 2Fe(OH)3 → Fe2O3 + 3H2O

3. The Water Electrode: In oxidizing acids the cathode reaction produces water as abyproduct:

O2 + 4H+ + 4e - → 2 H2O

If a continuous supply of both oxygen and hydrogen is available, the water electrodeproduces neither a buildup of rust nor a high concentration or dilution of ions at thecathode.

33

When a metal is exposed to an aqueous environment it develops an electric potential called a half-cell potential that is characteristic of the metal and the environment. This potential can bemeasured by using a reference electrode. Table 12.1 from Smith lists the half-cell potentials understandard conditions, frequently called emf series, of several metals against a hydrogen referenceelectrode. If two dissimilar metals are connected an electron flow occurs due to their potentialdifference. The metal with lower potential is the anode and dissolves to produce ions andelectrons while the metal/alloy with a higher potential is the cathode, which accepts electrons viametal circuit. Note that electrons move only through external circuits and ions move throughliquid solutions or electrolytes.

from W.F. Smith, Principles of Materials Science and Engineering

34

Protection Against Electrochemical Corrosion

A number of techniques are used to combat corrosion, including design, coating, inhibitors,cathodic protection, passivation, and material selection.

1. Coatings

Coatings are used to isolate the anode and the cathode regions. The category of coatingsincludes oil and paint type applications as well as electroplating. The electroplating processconsists of immersing a part to be coated in a solution of the metal to be coated and passingdirect current between the part (cathode) and another electrode (anode). The finalcomposition and the nature of the plating depends on various factors like temperature,current density, time and composition of the bath. A majority of metals like copper,chromium, zinc, nickel, cadmium, etc. can be applied by using this process.

2. Inhibitors

An inhibitor is a substance which when added in small quantities to a corrosive fluiddecreases the corrosion rate by several orders of magnitude. An inhibitor can be regardedas a retarding catalyst. They can be classified into several groups according to themechanisms by which they function.

3. Cathodic Protection

Cathodic Protection is a common method of corrosion prevention employed in pipelineswhere the metal behaves entirely as a cathode and hence corrosion is averted. This can beachieved in two ways:

a. Impressed Voltage: An external dc power supply is connected to the metal to beprotected. The negative terminal of the power supply is connected to the metaland the positive terminal to an inert anode like graphite. The process can bedescribed as the passage of a suitable electric current in an opposite direction at anequal or a greater rate than it was flowing when the metal was not protected.

b. Sacrificial Anode or Galvanic: A metal which has a higher tendency to corrode,anodic from the emf series, is coupled with the metal to be protected. Thesacrificial anode corrodes, supplies electrons to the metal, and thereby prevents ananode reaction at the metal. The sacrificial anode is consumed and must eventuallybe replaced.

35

4. Passivation

Certain metals protect themselves by forming an ultra thin (<100A thick) surface layer ofan oxide which forms a shield to oxygen penetration, e.g. stainless steel forms a thin layerapproximately 20A of chromium oxide. The metal is said to passivate itself in these cases.Passive films can be formed on brasses, aluminum alloys etc. by the use of oxidizersadded to solution. For example, chromate and nitrate salts are passivators.

Procedure

1.0 Half-Cell Potentials:

In this part of the experiment, the electrode potentials of a number of metals and alloys will bemeasured and a galvanic series will be constructed. The electrolyte used is 3.5wt% NaCl.

Clean the materials (see table below) by buffing them with emery paper followed by swabbingwith Fantastik or other cleaner using cotton balls and a final rinse in water. The metal is cleanwhen the rinse water “sheets”, i.e. does not ball up on the surface. Keep fingers off the surfaces,as oils in the skin will contaminate the surface. All specimens should be cleaned before a newtest is performed. Place quickly into the NaCl once clean without drying.

Measure the half-cell potentials with respect to the reference electrode by using a voltmeter. Avoltmeter has a high impedance and prevents current from flowing between the two electrodesand thus a stable reading of voltage gives the difference between the electrode potentials of themetal and the reference electrode under effectively open circuit conditions. It will probably takea few minutes for the voltage reading to settle to an equilibrium value. If it does not settle to asteady-state value, this is often due to inadequate cleaning. Construct the galvanic series for thefollowing materials:

Material Electrode Potential

Reference ElectrodeBrass

1090 steelAluminum alloy

ZincTitanium alloy

Ni alloyStainless steel

36

2.0 Potential Difference:

Two different metals/alloys are connected and the potential difference between the two ismeasured with a voltmeter. The measured potential difference should be theoretically equal to thedifference between the half-cell potentials measured for the metals/alloys individually. Calculatethe expected potential difference between the two metals using the formula:

Potential Difference = Ecathode – Eanode

Where Ecathode = the half cell potential of the cathodic material; and

Eanode = the half cell potential of the anodic material.

Measure the potential difference for the galvanic couples listed below and indicate which is theanodic material. You can also monitor the galvanic current, which can be used as an indication ofthe rate of galvanic corrosion because it represents electrons flowing from the anodic material tothe cathodic material. The current can be monitored with the same voltmeter withoutdisconnecting it from the electrodes by bridging between the two terminals with a 100 ohmresistor, the voltage across the resistor will allow estimation of the current flowing between theelectrodes as the voltmeter itself has a very high resistance and so all currently effectively flowsthrough the resistor.

Materials Theoretical Calculated Measured

Aluminum alloy – 1090 steel

1090 steel – Zinc

Brass –1090 steel

Titanium alloy – Brass

1090 steel – nickel alloy

3.0 Electroplating

Electroplating is often used as a means to provide corrosion protection to metals or to applydecorative/functional coatings . In this experiment you will electroplate copper onto steel. Thisis done for example as the first step as an underlayer to achieve adhesion for both nickel andchromium plating. The latter is used in many decorative applications such as on someautomobile wheels. Chrome plating is in fact a triple-layer plated coating with copper then nickelthen chromium on the top. The nickel provides the corrosion protection. The chromium is hardand shiny to make it look good and resist scratches.

Clean a piece of steel and one of copper as done previously. Use immediately without dryingand keep fingers or other contamination off of the cleaned surfaces. The two metals are to beconnected to the terminals of the power supply using the connectors provided. In order to

37

electroplate you will essentially be forcing the copper to corrode (it will become the anode) andthe positive copper ions that are formed in solution will then pass through the solution(electrolyte) and will be reduced (gain electrons) at the steel surface, which will be the cathode.Think about this and you should be able to decide to which terminal (positive or negative) eachmetal must be connected. Remember electrons flow in the opposite direction to conventionalcurrent in the external circuit through the power supply and ammeter. Use the ammeter tomonitor the plating current. Try to use a current that will give approximately 10-20 mA/cm2

current density based on the size of the steel electrode. Record the current and calculate thecurrent density. Also measure the plating time. To start plating immerse both electrodes in theelectrolyte WITH THE POWER SUPPLY TURNED ON. Plate long enough to achieve a coatingthickness of 10 micrometers. This can be calculated from Faraday’ Law assuming that all thecurrent is used to deposit copper ions on the steel.

Faraday’s Law states that one gram equivalent weight of metal will be deposited by the passageof 96,500 coulombs of charge. Coulombs = amps X seconds. Gram equivalent weight is gramatomic weight divided by valency. For copper the valency can be considered as 2, i.e. to depositone atom of copper it is necessary to provide 2 electrons to each copper ion.

To calculate the amount of current needed you need to calculate the total mass of copper to bedeposited. So determine the total area being plated, times the desired thickness gives the volumeof copper to be plated. Knowing the density of copper you can calculate the mass to bedeposited. Use Faraday’ Law to convert the desired mass of copper into the total amount ofcurrent that must be passed. This is how it is done in industry to this day. In most practicalplating some of the current gets used in other non-plating reactions, such as hydrogen generation,so the efficiency of current conversion to plating has to be known to do the calculation. Forcopper it is close to 100%. For other metals it might be quite low.

Questions:

1. Why is a stainless steel "stainless?"

2. What is the purpose of "galvanizing" a steel with a surface coating of zinc?

38

MODULE III – MATERIALS PROCESSING

III.1 MELTING AND CASTING

Objective

The purpose of this experiment is to introduce the process of casting and to evaluate thestrengthening effects of alloying. The system in question is aluminum-silicon.

Supplemental Reading

SMITH 6.1 (PP. 249-251) - CASTING OF METALS AND ALLOYS9.5 (PP. 538-541) - ALUMINUM CASTING ALLOYS8.6 - BINARY EUTECTIC ALLOY SYSTEMS

OHRING 8.2 - SOLIDIFICATION PROCESSING OF METALS5.5.1 - THE BINARY EUTECTIC PHASE DIAGRAM

Background

The casting and working of metals is as old as civilization itself. Today the quality andproperties of castings can be altered through careful control of alloy composition, melttemperature and casting conditions. During melting, alloys are brought to temperatures above themelting point of the system. Pouring temperatures are determined by the desire to havesufficient metal fluidity, fine cast grain structure and absence of excessive non-metallic phases(i.e., oxides, slags, etc.). In the laboratory, melting will be done via an induction furnace. Thefurnace consists of a motor generator set, operating at 10,000 cps and capable of delivering amaximum power of 20 kW. During operation, a water-cooled primary coil which couples to agraphite crucible-metal charge conductor combination (secondary), induces large AC currents inthe latter which in turn heats and melts the metal charge.

Once the metal is in the liquid state and at the desired pouring temperature, the metal ispoured into a waiting mold(s). There are numerous casting configurations, the following are just afew: continuous, static, lost wax, investment, centrifugal, die, slush, low pressure, sand, etc. Inall casting techniques, the melt superheat and evolved heat of fusion are removed through thecool mold walls upon solidification. The heat flow is such that the solidification front growsparabolically with time into the melt. On a microscopic scale, the solidification front is far fromuniform and consists of dendrite arrays in which there exists variations in composition. Thedendrites trap gas and contain porosity due to local solidification shrinkage. The grain structurewhich finally emerges is the result of complex nucleation, growth, heat transfer and fluidtransport effects. Castings are therefore structurally inhomogeneous and posses relatively lowductility.

39

Procedure

1. Material Preparation for Pure Al, Al-2wt, 4wt%, and 6wt% Si Alloys

1.1 Weigh the correct weight of material and fill the corresponding pre-marked crucibles.

2. Melting and Casting Procedure

2.1 Place crucible in the furnace. The teaching assistant will start the furnace.

2.2 ALWAYS USE SAFETY GLASSES WHEN NEAR THE FURNACE

2.3 Once the material is in the liquid state, skim the surface to remove unwanted oxides and pour into a metal mold. This step requires two participants.

2.4 After solidification, quench the mold and remove the sample.

2.5 DRY THE MOLD and complete steps 2.1 through 2.5 for the remaining alloys.

3. Hardness Values (Rockwell Hardness H-Scale)

3.1 For each sample, take five hardness values.

3.2 Plot the average hardness as a function of wt % Si.

Questions

1. What are the advantages of casting relative to other fabrication techniques (i.e., forging, rolling, extrusion, powder metallurgy, welding, etc.)?

2. How does the addition of Si affect the hardness of pure Al and by what mechanism does this affect occur? Hint: Use the Aluminum-Silicon phase diagram in your discussion.

3. How does mold design and material play a role in the mechanical integrity of the final part?

40

III.2 POWDER METALLURGY

Objective

The purpose of this lab experiment is to familiarize the student with a metal processing techniquewhich utilizes metal powder as the raw material. In particular, the effects of sinteringtemperature and time on the densification and mechanical properties of a brass alloy will beaddressed. Both the activation energy and the sintering mechanism(s) for this alloy will bedetermined.

Supplemental Reading

SMITH SMITH DOES NOT DISCUSS POWDER METALLURGY OR RELATED PROCESSES SUCH AS SINTERING. REFER TOOHRING OR OTHER INTRODUCTORY MATERIALS TEXT.

OHRING 8.4 - POWDER METALLURGY

SEE ALSO: www.mpif.org (THE METAL POWDER INDUSTRIES FEDERATION).

Background

A widely employed commercial method for fabricating a variety of metal and ceramiccomponents is based on the use of powders. The powders are blended to a requiredcomposition, compacted in dies of the desired shape and sintered at temperatures below themelting point of the system in question. The result is a solid part containing a residual amount ofporosity which is dependent on, among other things, the initial compact pressure, the sinteringtemperature and the sintering time. Generally, the extent of this residual porosity , or the extentof densification, can be modeled using the following Arhennius type equation.

∆V/Vo = A⋅ e-Q/RT⋅ tP …(1)

where ∆V is the change in volume due to sintering. Vo is the initial volume of the part. T is the sintering temperature (K). t is the sintering time (sec.).

R is the gas constant (cal/mol⋅ K). Q, P and A are constants.

Note that Q, P and A are constants, dependent on the mechanism by which mass isredistributed during the sintering process. Q and P represent the activation energy and the

41

sintering parameter of the system, respectively. Each will be calculated in this experiment.Examples of such mechanisms and their corresponding theoretical sintering parameters are listedbelow.

Sintering Mechanism P

Viscous Flow 1.00 Evaporation Condensation 0.67 Volume Diffusion 0.40 Surface Diffusion 0.29

Procedure

1. Press 9 Compacted Bar Specimens (Specimen Size 1.250” x 0.500” x 0.25”)

1.1 Weigh 18 grams of brass powder of composition 85% Cu - 15% Zn.

1.2 Pour the powder into the die cavity and shake to insure the powder is settled.

1.3 Place the die into the hydraulic press and apply 30,000 pounds of force.

1.4 Eject the specimen from the die cavity.

2. Sintering Sequence

2.1 After compacting, the length of each sample is measured with a bench micrometer and recorded.

2.2 The samples are then subjected to the following sintering sequence. Cover all samples with charcoal to retard oxidation.

Time/Temp. 750°C 850°C 950°C

30 min. Sample 1-1 Sample 1-2 Sample 1-3 60 min. Sample 2-1 Sample 2-2 Sample 2-3 90 min. Sample 3-1 Sample 3-2 Sample 3-3

2.3 After removal from the furnace, the final lengths are measured and recorded.

42

3. Determine the Activation Energy and the Sintering Parameter(s)

3.1 Using the approximation that 3⋅ ∆ l/lo ≈ ∆V/Vo we can rewrite equation (1) as:

∆l/lo = (1/3)⋅ A⋅ e-Q/RT⋅ …(2)

Equation (2) can be manipulated further by taking the natural log of both sides andfixing temperature and time, see equations (3) and (4).

ln(∆l/lo) = ln((1/3)⋅ A⋅ e-Q/RT) + P⋅ ln(t) (at fixed T) …(3)

ln(∆l/lo) = ln((1/3)⋅ A⋅ tP) − Q/R⋅ (1/T) (at fixed t) …(4)

3.2 Plot the ln(∆l/lo) vs. 1/T and the ln(∆l/lo) vs. ln(t). The slopes of each line

corresponds to Q/R and P, respectively. Use R = 1.987cal/mol⋅ K.

4. Measure the Transverse Rupture Strength (TRS) of the Specimens in Accordance with ASTM Standard B-528-70

4.1 Measure the load to failure (Pfail) using a special 3-point bending apparatus attached to anInstron tensile machine. The TRS is given by the following equation...

TRS =(3⋅ Pfail⋅ L)/(2⋅ t2⋅ w) …(5)

where Pfail is the load required for failure in lb. L is 1 inch. t is the specimen thickness (0.25”) w is the width of the specimen (0.50”)

P

(0.5⋅ P) L (0.5⋅ P)

43

4.2 Plot TRS vs. time and TRS and temperature. Note which processing variable has a greater effect on TRS and relate any observed trend to the densification of the sample.

Questions

1. What trends are observed between the TRS and the processing? Explain in a thorough manner.

2. What is the thermodynamic driving force which occurs during the sintering process?

3. What are some advantages and disadvantages of a powder process compared with other fabrication techniques?

44

III.3 THE PROCESSING AND PROPERTIES OF FERRITE MAGNETS

Objective

The purpose of this lab experiment is to expose the student to the fabrication of ferritematerial and to study the effect certain processing variables (i.e., sintering temperature and time)have on magnetic properties. In particular, the permeability, coercive field and the hysterisisbehavior of soft and hard magnets will be addressed.

Supplemental Reading

SMITH 11.2 - TYPES OF MAGNETISM11.4 - FERROMAGNETIC DOMAINS11.6 - THE MAGNETIZATION AND DEMAGNETIZATION OF A FERROMAGNETIC METAL

OHRING 14.3 - ATOMIC BASIS OF MAGNETISM14.4 - THE MAGNETIZATION PROCESS14.5 - FERROMAGNETIC MATERIALS

Background

One of the most significant recent advances made in ferromagnetic materials has been theintroduction and use of ceramic magnetic oxides (ferrites). Although the common ferritelodestone has been known and used for several millennia, it is only in the last 30 years thatcertain applications have required cheap insulating magnets, capable of high-frequency operationwith minimal eddy current losses . Ferrites are ferri magnets composed of oxides of transitionmetals that have the complex spinel or inverse spinel structure and a chemical compositioncorresponding to M+2Fe2

+3O4-2 where M represents a divalent cation, i.e., Ni+2, Fe+2,Cu+2, Zn+2,

Co+2, Sr+2 or Ba+2. Mixed components, e.g., (Ni, Zn) Fe2O4 also exist. In addition to thegenerally soft spinel ferrite magnets, there are the hard hexagonal ferrites of compositionMFe12O19 where M is typically Sr or Ba. An extensive discussion of structures, compositions,theory of sublattice magnetization and properties of ferrites can be found in the below references.

Procedure

1. Fabrication of Magnets

1.1 Weigh 8 grams of powder and pour into the die cavity.

1.2 Place the die set in the hydraulic press and apply 30,000 pounds of force. After pressing, eject the part.

45

1.3 Toroids and bars of soft material (Ni-ferrite) and bars of hard material (Sr-ferrite) will be produced following steps 1.1 and 1.2. A total of two toroids and one bar of soft and hard material will be pressed.

1.4 The samples are then subjected to the following sintering sequence (sintering time for all samples is approximately 45 min.):

Sample Sintering Temperature(°C)

Bar (hard material application) 950Bar (soft material application) 950Toroid (soft material application) 750Toroid (soft material application) 950

1.5 After the allotted sintering time, remove all samples from the furnaces and allow to air cool.2. Determination of Coercive Field (Difference Between Soft and Hard Magnet)

2.1 The hard and soft ferrite bars are to be magnetized by placing them between the pole faces of an electromagnet capable of developing a field strength in excess of 5000 gauses (see schematic of circuitry below). Magnetize specimens to a current of about 1.5 amps.

P.S.

RΩ EH

2.2 Next reduce the electromagnet current to zero and then REVERSE magnet current polarity. Very slowly raise the current. Look and listen for evidence of a change in magnetization of the specimen (i.e., a physical reorientation of the bar, an audible click, etc.) The critical value of the magnetizing current across a fixed resistor is a relative measure of the coercive field. Actual coercive fields can be obtained from known field strength behavior of the electromagnet as a function of current.

46

3. Determination of Permeability From Initial B-H Curve

3.1 Wind 50 turns of fine bell wire around the sintered toriods and connect the toroid inductor to the indicated circuit (see schematic of the circuit below).

3.2 Vary an AC current (1000 cps) through the toroid and measure, simultaneously, the corresponding voltage (EH) across the series resistor (RΩ) and the voltage (EB) across the toroid. The two measured voltages are directly proportional to H and B, respectively. The following equations are necessary for the conversions.

H = (4⋅ √ 2⋅ EH)/(10⋅ L⋅ RΩ) …(1)

where H is the applied field (oersteds). L is the mean magnetic path length in (cm). RΩ is the 1Ω resistor.

B = (108⋅ EB)/(4.44⋅ f⋅ N⋅ A) …(2)

where EB is a function of both applied and magnetic field created by the material (gauss). f is the AC current signal frequency (1000 cps). N is the number of wire turns (50).

A is the cross-sectional area of the toroid (cm2).

Questions

1. What is the significance of the permeability and coercive field of a magnet?

2. Quantitatively show the difference between a soft and hard magnet and give applications for both.

3. Explain the quality of your magnets as a function of processing variables.

47

MODULE IV- MATERIALS MICROSTRUCTURE

IV.1 HEAT TREATMENT OF STEEL

Objective

The purpose of this lab experiment is to study the effects of various heat treatments andquenching procedures on the microstructure of a 1080 plain carbon steel (0.8 % carbon). Theresults of this experiment are to be related to a Time-Temperature-Transformation curve (TTTcurve) for each steel sample. In addition, the grain structure of Armco iron samples (99.97 % Fe)which have been deformed and annealed in various ways will also be studied.

Supplemental Reading

SMITH: 9.2 - THE IRON-IRON CARBIDE PHASE DIAGRAM 9.3 - HEAT TREATMENT OF PLAIN CARBON STEELS

OHRING: 5.5.4 - Fe-Fe3C PHASE DIAGRAM9.2.1 - 9.2.5 HEAT TREATMENT

Background

Steels are our most important engineering material. Without them, the type of machineryand tools required to establish any industrial activity would be difficult to imagine. A veryimportant property of steel is the ability to alter its hardness by simple heat treatments. Thehardened steel is capable of cutting and shaping other softer materials such as other steels,nonferrous materials, plastics, wood, stone, etc. This hardening treatment critically depends onthe rate at which the steel is cooled from high austenitizing temperatures. The Time-Temperature-Transformation curve was developed as a convenient way to describe the resultantstructure and phase make up of the treated steel as a function of both temperature and time.Unlike the TTT curve, equilibrium phase diagrams are not a function of time and are oftensuperseded by the TTT curve for this reason. Below is a TTT curve for 1080 steel with severalheat treatment procedures relevant to this experiment.

48

Procedure

1. The Following Previously Heat Treated Steels Will be Supplied by the TA:

Set A: Austenitized at 899°C for 1/2 hour, quenched in lead at 677°C, held there for the statedtime (see chart below) and water quenched to room temperature.

Specimen Time (sec.) %Pearlite

1 10 2 100 3 240 4 360 5 540 6 1,000 7 10,000 8 100,000

Set B: Austenitized as above, but quenched in lead at 649°C for the times shown below.

Specimen Time (sec.) % Pearlite Specimen Time (sec.) % Pearlite

1 5 4 20 2 10 5 30 3 15 6 60

Set C: Austenitized at 900°C for 1 hour, quenched to Tm, held for 30 seconds, re-heated to

315°C for 40 seconds and finally quenched to room temperature.

Specimen Quench Temp. Tm (°C)

Hardness RC

% Temp. Martensite

1 218 61 2 204 58 3 177 57 4 163 ---- 5 149 54 6 121 52 7 93 48

49

2. Metallography

2.1 The samples require a final polishing. The teaching assistant will assist in this procedure.

2.2 After polishing, the samples are etched with 1% Nital (nitric Acid + Alcohol) so as to reveal the different phases in the optical microscope.

3. Estimation of Percentage of Nonequilibrium Phases

3.1 Using an optical microscope, estimate the percentage of high temperature transformation product (pearlite) for the sample sets A and B, both of which are composed of SAE 1080 steels subjected to the corresponding heat treatment and quenching procedures stated in Step 1.

3.2 Specimens from set C were heat treated in such a way as to reveal the temperature at which martensite first begins to form (Ms). The specimens show a tempered and regular martensite mixture which can be observed as dark and light regions in the microscope. The former is produced from austenite if Tm<Ms and the 315°C treatment tempers martensite but does not affect any metastable austenite which may be present. If Tm > Ms, only untempered light etching martensite will form. Using the optical microscope, estimate the percent of tempered martensite in each sample.

4. Sketch the Microstructures

4.1 Make one sketch from each set of samples, noting the different amounts of each phase. A sketch sheet is provided at the end of this lab.

5. The Following Armco Iron Samples Will be Provided by the T.A.

Sample # History or Treatment Brinell Hardness

1 Hot rolled and annealed at 954°C for 2 hr. 81

2 Cold rolled (60% red), heat treated like #1 167 3 Rolled like #2, annealed for 1 hr. at 510°C 137

4 Rolled like #2, annealed for 10 hr. at 510°C 110

5 Rolled like #2, annealed for 100 hr. at 510°C 99

6 Rolled like #2, annealed for 1 hr. at 677°C 103

7 Rolled like #2, annealed for 1 hr. at 899°C 97

50

6. Follow the Sample Preparation as Described in Step 2 for Each Iron Sample

7. Sketch the Microstructure of Each Sample

Questions

1. By examining the Armco samples 3 through 7, how would you rate the relative effectiveness of time and temperature in producing the observed changes?

2. Does your visual estimations of the percentages of non-equilibrium phases for each 1080 steel sample correlate with that predicted by the TTT curve?

51

METALLOGRAPHY OF STEEL SPECIMENS

SPECIMEN, TREATMENT, COMMENTS

SPECIMEN, TREATMENT, COMMENTS

SPECIMEN, TREATMENT, COMMENTS

52

METALLOGRAPHY OF ARMCO IRON SPECIMENS

Sample 1 Sample 2 Sample 3

Sample 4 Sample 5 Sample 6

Sample 7

53

IV.2 DIGITAL MICROSCOPY

Objective

The purpose of this lab experiment is to introduce digital image analysis as a viable meansto quantify the simply eutectic microstructures of several Bi-Cd and Sn-Zn alloy compositions.Conducting the experiment, the student will learn both eutectic systems and how to analyzephase amounts by employing a quantitative microscopy method. The data obtained will be usedto experimentally verify the Lever rule.

Supplemental Reading

SMITH: 8.4 - THE LEVER RULE 8.5 - NONEQUILBRIUM SOLIDIFICATION OF ALLOYS

8.6 - BINARY EUTECTIC ALLOY SYSTEMS

OHRING: 5.4 - INTRODUCTION TO BINARY PHASE DIAGRAMS5.5 - ADDITIONAL PHASE DIAGRAMS

Background

The quantitative metallography system used in this experiment consists of a high qualityreflection metallurgical microscope interfaced to a personal computer. Video images are digitizedby a frame grabber board inside the pc. Frame grabbers are now rather common. They are, forexample, often used by professional photographers to offer their clients "instant" proofs. Theboards associated with the Stevens laboratory are made by a company called Scion( www.scion.com) . Once in digital form, image data can be subjected to a large variety of image-processing algorithms. Many different software packages are available for image processing.They are used in applications ranging from military night vision to character recognition. TheStevens Design V lab uses software written at the National Institutes of Health (NIH) called NIHImage. Information and a download (Mac) can be obtained at http://rsb.info.nih.gov/nih-image/.Scion Corp. wrote a Windows-compatible version ( www.scion.corp) . The NIH site also refers to aJava-based version available at http://rsb.info.nih.gov/ij/ .

When eutectic alloys cool, two phases can be distinguished under the microscope. Bypolishing and chemically etching the samples, the two phases appear either dark or light in anoptical microscope. The phase proportions are derived from the phase diagram and the leverrule. The fractional amount of proeutectic phase or two-phase eutectic mixtures in the image arerelated to the phase diagram and can be determined optically. The digital image analysis facilityalso enables a count of the number of separate particles in the field of view as well as a statisticaldetermination of their areas, perimeters, shapes, etc. We will be using metallographic specimensmade from Bi-Cd and Sn-Zn alloys to determine the area percentage of a complex two-phaseimage. The relevant phase diagrams are given below.

54

The diagrams below are from: Metals Handbook 8th ed, Vol. 8, Metallography, Structures, andPhase Diagrams, published by the American Society for Metals (www.asminternational.org).

Procedure

1. Specimen History

1.1 The alloys you will be observing were prepared by carefully weighing known amounts of Bi and Cd or Sn and Zn. The material was melted and slowly cooled so as to maintain the equilibrium structure predicted by the phase diagram. After solidification, the samples were polished and etched in 1% FeCl3 solution.

1.2 Note - the metal surfaces are very soft and easily scratched. Please take car inhandling them. If you are told to re-polish the specimens, use a clean abrasive- free cloth andwet polish. The specific compositions are listed in tables 1 and 2 at the end of this lab section.

55

2. Observe the Specimens Through an Optical Microscope

In small groups, image your specimen on the high-resolution monitor. Identify the proeutecticphase and eutectic mixture in each sample. Estimate, by eye, the relative amounts of each phasepresent. Sketch two microstructures, one of a hypoeutectic alloy and one of a hypereutecticalloy in either the Sn-Zn of Bi-Cd system.

3. Imaging of the Specimens

Analyze the amounts of each phase in the image by the following method and record in theappropriate tables provided in this lab section. The TA will outline the relevant menus andcommands needed to use the Image software. Note that there is a useful online help menu. A. Freeze the image by grabbing the frame in the computer. B. Binarize the image to render it fully black and white. C. Take three measurements from three different regions of the sample.

4. Lever-rule calculation

For each composition studied, calculate the amount of proeutectic and eutectic structure byapplying the lever rule to the Sn-Zn and Bi-Cd phase diagrams. Plot experimental and calculatedfractions of proeutectic structure versus composition for each alloy system.

5. Quantifying characteristic length scales

Use a standard specimen with microstructural features with a known length scale to calibrate themagnification associated with a digital image. The TA will provide copper TEM grids (3mmdiameter copper screens used to support specimen for study in a transmission electronmicroscope). The TA will outline the appropriate software functions (e.g. SetScale). Once thescale is calibrated for a given objective lens on the microscope, images of alloy specimens can bestudied using the same objective lens. Us the measuring functions on the software toquantitatively estimate the length scale characteristic of each phase in the pro-eutectic andeutectic structures.

Questions

1. How well is the Lever rule obeyed in these alloy systems?

2. What corrections would have to be made to account for differences in density between thecomponents or phases?

3. What errors are involved in the image analysis system? From this, estimate the error in thecomputer calculation of the phase proportion.

56

4. Why is the length scale associated with the pro-eutectic phase substantially greater than thatassociated with the the same phase in the eutectic structure?

5. How is contrast between different microstructural features generated in the opticalmicroscope?

6. How does the length scale associated with different microstructural features affect theproperties of the alloy?

Table 1 Sn-Zn Alloys

Sn - Zn Alloys Group A Group A Group B Group B Composition Eye Cue 2 Eye Cue 295 Sn - 5 Zn

93 Sn - 7 Zn

91 Sn - 9 Zn

60 Sn - 40 Zn

70 Sn - 30 Zn

Table 2. Cd-Bi Alloys

Cd - Bi Alloys Group A Group A Group B Group B Composition Eye Cue 2 Eye Cue 280 Cd - 20 Bi

70 Cd - 30 Bi

60 Cd - 40 Bi

40 Cd - 30 Bi

30 Cd - 70 Bi

20 Cd - 80 Bi

10 Cd - 90 Bi

57

IV.3 X-RAY ENERGY DISPERSIVE ANALYSIS OF MATERIALS

Objective

This experiment uses a method based on characteristic X-rays to determine the types of elementsin unknown specimens as well as quantify their compositions.

Supplemental Reading

SMITH: 2.3 - THE ELECTRONIC STRUCTURE OF ATOMS 3.11 - X-RAY SOURCES (PP. 103-105)

OHRING: 2.2 - ATOMIC ELECTRONS IN SINGLE ATOMS (FOCUS ON 2.2.4)2.3 - FINGERPRINTING ATOMS (ESPECIALLY 2.3.1)

Background

Two important ways to characterize materials are through microscopic observation and chemicalanalysis. X-ray energy dispersive (EDX) analysis is a means of performing rapid chemicalanalysis. EDX analysis is quick and can be very accurate.

The physical phenomenon which makes EDX analysis possible is based on the emission anddetection of characteristic X-rays. If impingement of high-energy particles (e.g. electrons orgamma rays) on an atom ejects K shell electrons, and electrons from L shells fall into the K shellvacancy, a photon in the form of a K X-ray is emitted. Similarly, M to L shell processesgenerate L X-rays, etc. The combination of all X-rays generated by an atom is its X-rayspectrum. The X-ray photon generated by a transition has an associated wavelength (λ), so anenergy value can be assigned because of the deBroglie relationship:

E = h⋅ ν = h⋅ ( c/λ)

where ν frequency c speed of light h Boltzman’s constant

The energy of the photon produced by each K, L, M,... transition is different for each element.Thus, these photons are called characteristic X-rays. The combinattion of all X-rays generatedby an atom or a material constitutes its X-ray spectrum. The amount of a particular element canbe determined by quantifying the number of characteristic X-rays it emits and compare this tosome standard. Typical energies of the characteristic K X-rays range from 0.05 KeV for Li up to97 KeV for U. The L and M X-rays have lower energies.

58

The EDX system used in this laboratory consists of three main components, an excitationsource, a detection readout system, and the specimen in question (see schematic below). Thesource is a radioactive isotope of Am 241 which emits gamma rays used in the excitation of atomicelectrons from the sample or unknown specimen. The detection-readout system is acryogenically cooled silicon detector doped with Li, i.e., Si(Li) detector. When a characteristic X-ray impinges on the detector it creates electron-hole pairs thus creating an integrated chargeproportional to the energy of the X-ray. This charge produces a voltage pulse which issubsequently amplified. The magnitude of each voltage pulse is proportional to the energy of theX-ray incident on the Si(Li) detector.

Using a Multi-channel analyzer (MCA), the computer breaks the amplified signal into 1024voltage ranges or "channels." Each channel records the number of pulses in its voltage rang, andthe contents of the channels are displayed on the monitor as an X-ray spectrum. The verticalscale is the number of pulses, and the horizontal scale represents the 1024 channels. Thechannels can be calibrated to represent the X-ray energies. he X-rays generated by the sample inthe microscope are converted into voltage pulses by the detector, amplified, then fed into thecomputer marked "MCA". Each voltage pulse is proportional to the energy of an X-ray. Thecomputer breaks the amplified signal into 1024 voltage ranges, or "channels". Each channelrecords the number of pulses in its voltage range, and the contents of the channels are displayedon the screen as a spectrum. (The vertical scale is the number of pulses, the horizontal scalerepresents the 1024 channels.) The channels can be calibrated to represent the X-ray energiesthat were generated by these pulses, and the peaks in the spectrum will indicate the energies ofthe characteristic X-rays of the sample.

Computer w/ MCA card

Americium sourceLead (Pb) shield

Liquid nitrogen (LN at 77 K)

Amplifier

Pre -amp

Si(Li) X-ray detector

Sample

Cu thermal conductor

59

Procedure

1. Data Acquisition

1.1 Using X-ray energies from known samples, the energy scale will be calibrated.

1.2 The X-ray spectra from various unknown specimens will then be collected and displayedon Windows computers using EG&G Maestro software. The TA will demonstrate how to usethe necessary functions of this software.

1.3 The X-ray spectra for two known Zn-Sn alloy compositions will be collected. From thesedata, the compositions of three other Zn-Sn unknowns will be determined.

2. Data Interpretation

The TA will provide either a soft copy copy of a database entitled xraydata.mdb. This is in aformat readable by Microsoft Access on Windows. It lists the characteristic X-ray energies forvarious elements. Access enables flexible search capabilities. A portion is reproduced below.

atomic number element X-ray X-ray energy25 MN LB1 64925 MN LL 55625 MN LN 56725 MN LG5 64925 MN LB9 72125 MN LB6 64025 MN LB3 72125 MN LA2 63725 MN LA1 63725 MN KB3 648925 MN KB1 648925 MN KA2 588725 MN KA1 589825 MN LB4 72126 FE LB6 70826 FE LB1 71826 FE LL 61526 FE LN 62826 FE LG5 71826 FE LB9 79226 FE KA1 640326 FE KA2 639026 FE LA2 70526 FE LA1 70526 FE KB3 705726 FE KB1 705726 FE LB4 79226 FE LB3 792

60

Questions

1. How do EDX and X-ray diffraction methods differ?

2. What are some limitations to the use of EDX analysis method?

# of Elements Energies (keV) Intensities Elements

1 1 Kα = Kβ =

2 1 Kα = Kβ =

3 1 Kα = Kβ =

4

5

6

7

8

9

# wt% Composition Intensity (cm)

1 10 Zn - 90 Sn

2 90 Zn - 10 Sn

3

4

5

61

Spreadsheet analysis of EDS X-ray compositional data

The raw data produced by an EDS X-ray detection system amounts to a two-column arraywhere one column corresponds to a channel number and the other corresponds to the number ofX-ray counts in each channel. These data can be analyzed using a spreadsheet program such asExcel. There are two principal chores: (I) calibrate the energy scale using a known sample; and(ii) identify the elements in an unknown sample.

The Teaching Assistant will provide two soft data files. These are text files in a standardizedformat (see the Microscopy Society of America - MSA at www.microscopy.org). The files aretwo-column arrays with the channel number in the first column and the number of counts in thesecond column. There is header information at the top of the file. One file corresponds to asample whose composition is known (e.g. pure Fe). This sample will provide a combination ofpeaks in its X-ray spectrum which can be used for calibration of the energy scale. The other filecontains data from an unknown sample.

Calibration using the known standard

Open the data file of the known sample using a spreadsheet such as Excel. The software shouldprompt you to specify which line to start reading from (avoid any header lines) as well as thenature of the delimiters (commas in the present case). Chart the raw data (an x-y scatter plot isbest in Excel) in order to better visualize the data. The raw data should have 1024 channels.Depending on the electron energy used to excite the -rays in the specimen, the upper portion ofthe data set (channels > 500) may only contain a few counts per channel. These can be ignored inorder to facilitate analysis. The example given in figure 1 comes from iron (Fe). It has fourpeaks. The largest is near the origin circa channel 10. This peak is an artifact from the datacollection process. It should be ignored. The second peak is circa channel 40. It corresponds toan L excitation. The two peaks near channels 330 and 370 correspond to the Kα and Kβ

excitations, respectively. The energies of the Fe Lα, Kα, and Kβ peaks can be found in the X-ray

energy database (xraydata.mdb). The weighted Kα energy is 6.3989 keV. The Kβ energy is

7.0571 keV. The Lα excitation is the most intense of the L lines. Its energy is 0.7050 keV.

There are essentially two unknowns associated with the energy-scale calibration. One is the zeroand the other is the energy per channel. The energy associated with a specific channel can berepresented as a linear function with the form: Ei=mci+b where Ei and ci are the energy andnumber of the ith channel, respectively. B is an offset which fixes the zero. M is a scalingcoefficient to convert channel number to energy. The two unknowns, m and b, can be determinedknowing the energies associated with two specific peaks and writing two equations in twounknowns. In the present example, the channels corresponding to the maximum number ofcounts in the Ka and La peaks are 331 and 46, respectively. Less error is introduced by usingwidely separated peaks for the calibration (e.g. La and Ka rather than Ka and Kb). A moresophisticated and accurate method would involve modeling each X-ray peak with a Gaussian

62

Figure 1 - X-ray spectrum from pure Fe with an uncalibrated energy scale

function and then find the channel, in fractional units, corresponding to the center of eachGaussian. The present analysis gives:

m = (E2-E1)/(c2-c1) = (6.3989 - 0.7050)/(331-46) = 0.0200 keV/channel. b = E1 - mc1 = -0.2211 keV

These constants (b and m) can be used to create a new column in the spreadsheet where thechannel number is converted to X-ray energy. The result of this operation is illustrated in figure2. While the curve looks very similar to the previous figure, the energy scale is now fixed.

Identification of elements in an unknown specimen.

Using the values of m and b determined from the known specimen, the energy scale of a spectrumfrom an unknown specimen can now be calibrated. The energies associated with each peak canthen be identified using the searchable database to identify the most probable element or elementsassociated with each peak. Some judgement may have to be used here because of overlappingenergies from different elements.

0

2000

4000

6000

8000

10000

12000

14000

0 100 200 300 400 500

Channel Number

X-r

ay C

ounts

63

Figure 2 - X-ray spectrum from pure Fe after energy-scale calibration.

An example from an unknown specimen is given in fig. 3. It is from a precipitate in a steel. Thethree indicated peaks most likely correspond to Al Ka (1.487 keV), Si Ka (1.7395 keV), andeither S Ka (2.3065 keV) or Mo La1 (2.2931 keV). The third peak is most likely sulfur.Oxysulfide phases are often found in steels.

Figure 3 - Calibrated X-ray spectrum from precipitate particle of unknown composition.

0

2000

4000

6000

8000

10000

12000

14000

0 2 4 6 8 10

X-ray Energy (keV)

X-r

ay C

ounts

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 2 4 6 8 10 12

X-ray Energy (keV)

X-r

ay C

ounts

Series1

1.4789

1.73892.2989

64